FPGA-Based Sensorless PMSM Drive With Sliding Mode Observer for Velocity Tracking

This paper presents the design and implementation of sensorless permanent magnet synchronous motor (PMSM) drives on field programmable gate arrays (FPGAs) using a sliding mode observer (SMO) for applications requiring highly accurate velocity tracking. To eliminate mechanical position sensors while reducing system cost and vulnerability to harsh environments, a chattering-reduced SMO employing a hyperbolic tangent switching function is proposed. Particular attention is given to practical FPGA feasibility, avoiding excessive algorithmic complexity while ensuring precise tracking performance during speed variations. The proposed estimation scheme is validated through both simulation and real-time experimental studies on an FPGA platform, exploiting its inherent parallel processing capability for deterministic real-time execution. The results demonstrate that the proposed approach achieves high accuracy in rotor speed and position estimation while maintaining stability and robustness against parameter variations and external disturbances. This work represents a significant advancement toward efficient real-time control of PMSM drives, offering a low-cost and reliable solution for industrial applications with diverse velocity profiles. To quantitatively validate the proposed design, extensive evaluations were conducted under identical conditions. Compared with a standard SMO implemented on the same FPGA platform and tested under the same scenarios, the proposed method reduces the speed estimation error by 49.5% in mean absolute error (MAE) and by 47.2% in root mean square error (RMSE). In addition, the rotor flux-angle error is reduced by 4°– 5°, current chattering is suppressed from ±0.15Ato below ±0.05A, and the back-EMF total harmonic distortion (THD) is reduced by 1.82%. These improvements are achieved with only a modest increase in FPGA resource utilization, from 12.57% to 15.74%, verifying effectiveness of the proposed approach.